Superconductivity — Cooper pairs, BCS theory, and macroscopic quantum order

Intuition Beginner

A superconductor is a material that, below a sharply defined critical temperature, carries electric current with no resistance at all. In 1911 Heike Kamerlingh Onnes cooled mercury to 4.2 kelvin and watched its resistance vanish — once started, a current would run forever, losing no energy to heat.

Below the same temperature the material also expels magnetic fields from its interior, the Meissner effect. A magnet placed over a superconducting dish floats in mid-air, held up by the field it cannot push inside. This is not mere shielding; it is an active rearrangement of surface currents.

The explanation took nearly fifty years. In 1957 Bardeen, Cooper, and Schrieffer showed that at low temperature electrons bind loosely into pairs — Cooper pairs — by trading lattice vibrations (phonons). The pairs move in lockstep as one giant quantum object, and the lockstep makes scattering impossible, so resistance drops to zero.

Superconductivity is a phase of matter, like ice versus water. Cross the critical temperature and the electrons reorganise abruptly into the paired, ordered state. The deeper lesson is macroscopic quantum order: quantum mechanics acting on billions of particles at once, visible to the naked eye.

Visual Beginner

A resistance-versus-temperature curve: the metal's resistance falls slowly as it cools, then at $T_c$ drops vertically to exactly zero. Beside it, a magnet levitates above a superconducting pellet, held up by expelled magnetic flux.

Above $T_c$ the material is an ordinary metal. Below $T_c$ it is a superconductor: zero resistance and expelled magnetic flux, two signatures of a single new phase.

Worked example Beginner

Flux quantisation. In a superconducting ring, magnetic flux comes in fixed packets called flux quanta. The size of one packet is ${\Phi }_0=h/(2e)$, where $h=6.626\times 10^{-34}$ joule-second is Planck's constant and $e=1.602\times 10^{-19}$ coulomb is the electron charge. The factor of 2 appears because the carriers are Cooper pairs, each carrying charge $2e$.

Step 1. Multiply the charge by two: $2e=3.204\times 10^{-19}$ coulomb.

Step 2. Divide Planck's constant by this: ${\Phi }_0=6.626\times 10^{-34}/3.204\times 10^{-19}=2.07\times 10^{-15}$ weber.

Step 3. One flux quantum is about $2.07\times 10^{-15}$ weber. A superconducting loop traps flux only in whole multiples of this value — a direct, countable sign of the underlying quantum order.

Critical temperatures of a few superconductors:

Material	$T_c$ (K)	Year found
Mercury	4.2	1911
Niobium	9.2	1930
YBa${}_2$Cu${}_3$O${}_7$ (YBCO)	93	1987

Conventional superconductors need liquid helium, near 4 K. The copper-oxide YBCO works at the temperature of liquid nitrogen, 77 K — far cheaper to maintain, and the breakthrough that opened high-temperature superconductivity.

Check your understanding Beginner

Formal definition Intermediate+

A material is superconducting below its critical temperature $T_c$ when two properties coexist: perfect DC conductivity, $\rho =0$, and perfect diamagnetism, the Meissner effect. The transition at $T_c$ is a second-order phase transition (in zero applied field) at which a macroscopic quantum order parameter appears ^{[Tinkham 1996]}.

The Meissner effect states that for a simply-connected sample cooled below $T_c$ in any applied field, the magnetic induction vanishes in the interior: $\mathbf{B}(\mathbf{r})=0$ in the bulk. This holds independent of the path taken in the field–temperature plane and is not a consequence of zero resistance. The London brothers captured the electrodynamics by postulating, in the London gauge, ${\mathbf{j}}_s=-(n_s{e}^2/m)\mathbf{A}$, which yields ${\nabla }^2\mathbf{B}=\mathbf{B}/{\lambda }_L^2$ with penetration depth ${\lambda }_L^2=m/({\mu }_0{n}_s{e}^2)$, where $n_s$ is the superfluid density. Fields decay exponentially over ${\lambda }_L$ (typically $\sim 10^{-7}$ m).

The Ginzburg-Landau order parameter $\psi (\mathbf{r})=|\psi (\mathbf{r})|e^{i\theta (\mathbf{r})}$ is a complex macroscopic wavefunction with $|\psi {|}^2=n_s$. The coherence length $\xi$ sets the distance over which $\psi$ can vary; it diverges as $T\to T_c^{-}$. A magnetic field above the critical field $H_c(T)$ destroys superconductivity, with $H_c(T)=H_c(0)[1-(T/T_c{)}^2]$ near $T_c$.

Flux through any superconducting loop is quantised in units ${\Phi }_0=h/(2e)\approx 2.068\times 10^{-15}$ Wb. In type-I superconductors flux is expelled completely up to $H_c$ and the normal state then invades abruptly; in type-II superconductors, for which the GL parameter $\kappa ={\lambda }_L/\xi >1/\sqrt{2}$, flux penetrates between lower and upper critical fields $H_{c1}$ and $H_{c2}$ as quantised vortices, each carrying one ${\Phi }_0$.

Counterexamples to common slips

• Zero resistance is not the Meissner effect. A perfect conductor would trap whatever field was present at cooling; a superconductor expels it. The Meissner effect is what distinguishes superconductivity from a merely perfect metal.
• Superconductivity is not low-temperature metallic behaviour. It is a distinct thermodynamic phase with a jump in heat capacity at $T_c$, a hidden order parameter, and flux quantisation.
• The BCS gap is not a band gap. The energy gap $\Delta$ opens at the Fermi surface inside an otherwise metallic band, and its size (order 1 meV) is thousands of times smaller than a typical semiconducting band gap.

Core model Intermediate+

The Ginzburg-Landau (GL) free-energy functional is a phenomenological Landau theory of the superconducting transition, written in 1950 and justified microscopically only after BCS. For a superconductor in a magnetic field with vector potential $\mathbf{A}$,

$$F[\psi ,\mathbf{A}]=\int d^{3}r\left[\alpha |\psi {|}^2+\frac{\beta }{2}|\psi {|}^4+\frac{1}{2m^{\ast }}{\left|\left(-i\hslash \nabla -2e\mathbf{A}\right)\psi \right|}^2+\frac{|\mathbf{B}{|}^2}{2{\mu }_0}\right].$$

Here $\alpha ={\alpha }_0(T-T_c)$ changes sign at $T_c$ — the Landau signature of a second-order transition — $\beta >0$ stabilises the ordered phase, and $m^{\ast }=2m_e$ is the pair mass. The factor $2e$ in the covariant derivative reflects that the condensate carries charge $2e$, which already in 1950 hinted at pairing ^{[Ginzburg & Landau 1950]}.

Minimising $F$ over $\psi$ at $\mathbf{A}=0$ gives the bulk order parameter $|\psi {|}^2=-\alpha /\beta$ for $T<T_c$ and zero above. The GL equations obtained from $\delta F/\delta {\psi }^{\ast }=0$ and $\delta F/\delta \mathbf{A}=0$ reduce to the London equation deep in the bulk and introduce the GL parameter $\kappa ={\lambda }_L/\xi$: type-I materials have $\kappa <1/\sqrt{2}$, type-II have $\kappa >1/\sqrt{2}$.

GL theory predicts two characteristic lengths: the penetration depth ${\lambda }_L$ (the range over which fields are suppressed) and the coherence length $\xi$ (the stiffness of the order parameter). Gor'kov showed in 1959 that GL theory is the near-$T_c$ limit of BCS theory, with $\xi$, ${\lambda }_L$, and the coefficients $\alpha ,\beta$ expressible in terms of the microscopic gap $\Delta$ and the density of states.

Key derivation Intermediate+

The central microscopic fact is the Cooper instability: the normal Fermi sea of a metal is unstable to bound electron-pair formation for an arbitrarily weak attractive interaction. This resolves why superconductivity is a sharp phase transition rather than a gradual cooling effect, and why it appears in so many metals at low enough temperature.

Cooper problem. Consider two electrons just above a filled Fermi sea, with momenta $\mathbf{k}$ and $-\mathbf{k}$ (total momentum zero, opposite spins), interacting through a weak attraction $V$ within a thin shell $0<{\epsilon }_k-{\epsilon }_F<\hslash {\omega }_D$ whose width is set by the Debye frequency — the phonon-mediated attraction of 12.20.01. Treating the filled sea as a forbidden region, the pair amplitude $\varphi (\mathbf{k})$ satisfies

$$(E-2{\epsilon }_k)\varphi (\mathbf{k})=-V\sum _{{\mathbf{k}}^{\prime }}\varphi ({\mathbf{k}}^{\prime }),$$

with both $\mathbf{k}$ and ${\mathbf{k}}^{\prime }$ inside the Debye shell. Take $V$ constant and attractive. Setting $C={\sum }_{{\mathbf{k}}^{\prime }}\varphi ({\mathbf{k}}^{\prime })$ gives $\varphi (\mathbf{k})=VC/(2{\epsilon }_k-E)$; summing over $\mathbf{k}$ in the shell and cancelling $C\ne 0$,

$$1=V\sum _{\text{shell}}\frac{1}{2{\epsilon }_k-E}\approx VN(0){\int }_0^{\hslash {\omega }_D}\frac{d\epsilon }{2\epsilon +E_b},$$

where $N(0)$ is the single-spin density of states at the Fermi level and we wrote $E=2{\epsilon }_F-E_b$ for a bound state below the continuum. The integral is $\frac{1}{2}\ln \left(\frac{2\hslash {\omega }_D+E_b}{E_b}\right)$, giving the binding energy

$$E_b\approx 2\hslash {\omega }_D\exp \left(-\frac{2}{N(0)V}\right).$$

For any positive $V$, however small, $E_b>0$: a bound pair exists, and the Fermi sea is unstable.

BCS ground state and gap. Bardeen, Cooper, and Schrieffer promoted the Cooper pair into a many-body ground state, a coherent superposition of paired states ^{[Bardeen, Cooper & Schrieffer 1957]}:

$$|\mathrm{BCS}⟩=\prod _{\mathbf{k}}(u_{\mathbf{k}}+v_{\mathbf{k}}{c}_{\mathbf{k}\uparrow }^{†}{c}_{-\mathbf{k}\downarrow }^{†})|0⟩,|u_{\mathbf{k}}{|}^2+|v_{\mathbf{k}}{|}^2=1.$$

Minimising the expectation of the reduced Hamiltonian fixes $u_{\mathbf{k}},v_{\mathbf{k}}$ through the quasiparticle energy $E_{\mathbf{k}}=\sqrt{({\epsilon }_k-{\epsilon }_F{)}^2+{\Delta }^2}$, where the energy gap $\Delta$ obeys the self-consistency (gap) equation

$$1=VN(0){\int }_0^{\hslash {\omega }_D}\frac{d\epsilon }{\sqrt{{\epsilon }^2+{\Delta }^2}}\approx VN(0)\ln \frac{2\hslash {\omega }_D}{\Delta },$$

so at $T=0$,

$$\Delta (0)\approx 2\hslash {\omega }_D\exp \left(-\frac{1}{N(0)V}\right)=1.764k_B{T}_c.$$

The gap $\Delta$ is the energy needed to break a Cooper pair. It vanishes continuously at $T_c$ and removes all low-energy single-particle final states, so there is nothing for an electron to scatter into — the microscopic origin of zero DC resistance.

Bridge. The Cooper instability builds toward 12.13.01, where the BCS state is recognised as a coherent state in fermionic Fock space and the Bogoliubov rotation defines the quasiparticle basis, and appears again in 08.02.02, where the same spontaneous-U(1)-symmetry-breaking structure governs every Landau phase transition. The foundational reason superconductivity is a sharp phase transition is that an arbitrarily weak attraction collapses the Fermi sea; putting these together, this is exactly the bridge from the electron-phonon interaction of 12.20.01 to the laboratory facts of zero resistance and flux quantisation, and the pattern generalises to unconventional and high-$T_c$ superconductors whose pairing glue is no longer phononic.

Exercises Intermediate+

Lean formalization Intermediate+

lean_status: none. Superconductivity deploys second-quantised many-body states — the BCS variational wavefunction, the Bogoliubov quasiparticle rotation, and operator-valued creation and annihilation fields — whose formal apparatus is the subject of 12.13.01 (Fock spaces). Mathlib currently lacks second-quantised Hilbert spaces, variational many-body states, and a gauge-theoretic treatment of the order parameter. The Ginzburg-Landau free-energy functional and the London equations could in principle be stated once the function-space and gauge-field infrastructure exist, but the Cooper instability and the BCS gap equation require substantial new analysis machinery. The present unit's correctness gate is therefore derivation consistency and experimental evidence, not formal proof.

Advanced results Master

Josephson effect. If two superconductors are separated by a thin insulating barrier — a Josephson junction — Cooper pairs tunnel coherently. Brian Josephson predicted in 1962 that a DC supercurrent $J=J_c\sin \delta$ flows with no voltage for a phase difference $\delta$ across the junction, and that a fixed voltage $V$ drives an AC current at frequency $\nu =2eV/h$ ^{[Josephson 1962]}. The voltage-frequency relation now defines the SI volt; the DC effect underlies SQUID magnetometers sensitive enough to detect the magnetic fields of the human brain.

Flux quantisation and vortices. Single-valuedness of the macroscopic wavefunction forces flux through any superconducting loop to be an integer multiple of ${\Phi }_0=h/(2e)$. In type-II materials, magnetic flux enters between $H_{c1}$ and $H_{c2}$ as a regular lattice of vortices (the Abrikosov lattice), each carrying exactly one ${\Phi }_0$ and winding the condensate phase by $2\pi$ around a normal core. The vortex lattice and its pinning by defects control the technologically crucial critical current of high-field magnets.

Isotope effect. Replacing the ions of a conventional superconductor with heavier isotopes lowers the phonon frequencies and shifts $T_c$ as $T_c\propto M^{-1/2}$. This direct dependence on ionic mass is the fingerprint of phonon-mediated pairing and was a key clue that guided BCS.

The high-$T_c$ frontier. In 1986 Bednorz and Müller found superconductivity at 30 K in a copper-oxide ceramic; within a year $T_c$ exceeded 77 K in YBCO, and cuprates now reach about 130 K at ambient pressure ^{[Bednorz & Müller 1986]}. These materials violate several BCS signatures — the isotope effect is anomalous, the gap has $d$-wave symmetry, and the pairing scale is too high for phonons alone — and no accepted microscopic theory exists. Iron-based superconductors, nickelates, and hydrogen-rich compounds under megabar pressures (with $T_c$ approaching room temperature) extend the frontier and keep the search for a unifying mechanism open.

Synthesis. Superconductivity is the paradigm of emergent macroscopic quantum order: the Cooper instability builds toward 12.13.01, where the BCS state is a coherent state in fermionic Fock space and the Bogoliubov transformation defines the quasiparticle basis, and appears again in 08.02.02, where the same spontaneous-U(1)-symmetry-breaking order parameter governs every Landau phase transition. The foundational reason it is a phase is that an arbitrarily weak attraction collapses the Fermi sea; this is exactly the bridge from the electron-phonon interaction of 12.20.01 to laboratory-scale zero resistance and flux quantisation, and the pattern generalises to unconventional and high-$T_c$ superconductors whose pairing glue is no longer phononic. Putting these together, the central insight is that a single coherent macroscopic wavefunction, with amplitude $|\psi |$ and phase $\theta$, accounts simultaneously for zero resistance, the Meissner effect, flux quantisation, and the Josephson effect.

Full proof set Master

Proposition (Cooper binding energy). Two electrons at the Fermi surface interacting via a constant attractive potential $-V$ ($V>0$) within a shell $0<{\epsilon }_k-{\epsilon }_F<\hslash {\omega }_D$ form a bound state with binding energy

$$E_b\approx 2\hslash {\omega }_D\exp \left(-\frac{2}{N(0)V}\right),$$

where $N(0)$ is the single-spin density of states at the Fermi level.

Proof. Write the pair state as $|\Psi ⟩={\sum }_{\mathbf{k}}\varphi (\mathbf{k})c_{\mathbf{k}\uparrow }^{†}{c}_{-\mathbf{k}\downarrow }^{†}|F⟩$, where $|F⟩$ is the filled Fermi sea and the sum runs over the Debye shell. Acting with the Hamiltonian $H_0+V$, where $H_0$ contributes kinetic energy $2{\epsilon }_k$ and $V_{\mathbf{k}{\mathbf{k}}^{\prime }}=-V$ inside the shell, gives

$$(E-2{\epsilon }_k)\varphi (\mathbf{k})=-V\sum _{{\mathbf{k}}^{\prime }\in \text{shell}}\varphi ({\mathbf{k}}^{\prime }).$$

Set $C={\sum }_{{\mathbf{k}}^{\prime }}\varphi ({\mathbf{k}}^{\prime })$. Then $\varphi (\mathbf{k})=VC/(2{\epsilon }_k-E)$. Summing both sides over $\mathbf{k}$ in the shell and cancelling the nonzero factor $C$,

$$1=V\sum _{\mathbf{k}\in \text{shell}}\frac{1}{2{\epsilon }_k-E}=VN(0){\int }_0^{\hslash {\omega }_D}\frac{d\epsilon }{2({\epsilon }_F+\epsilon )-E}.$$

A bound state lies below the two-particle continuum, so write $E=2{\epsilon }_F-E_b$ with $E_b>0$. The denominator becomes $2\epsilon +E_b$, and the integral evaluates to

$$\frac{1}{VN(0)}=\frac{1}{2}\ln \left(\frac{2\hslash {\omega }_D+E_b}{E_b}\right).$$

Hence $2/[VN(0)]=\ln (1+2\hslash {\omega }_D/E_b)$, giving $E_b=2\hslash {\omega }_D/(e^{2/[VN(0)]}-1)$, and for weak coupling ($VN(0)\ll 1$) the binding energy reduces to $E_b\approx 2\hslash {\omega }_D{e}^{-2/[VN(0)]}$. Because the exponent is finite for any positive $V$, no matter how small, $E_b>0$ always: the pair is bound, and the normal Fermi sea is unstable. $\square$

Connections Master

• Condensed-matter foundations 12.20.01. Band structure, the Fermi surface, and phonons supply both the electron states that pair and the phonon-mediated attraction that binds Cooper pairs in conventional superconductors.

• Fock spaces and second quantisation 12.13.01. The BCS ground state is a coherent state in fermionic Fock space, and the Bogoliubov transformation to quasiparticles is a canonical rotation of that Fock-space basis.

• Spontaneous symmetry breaking 08.02.02. The superconducting order parameter $\psi =|\psi |e^{i\theta }$ breaks the global U(1) gauge symmetry exactly as in the Landau theory of phase transitions; its phase stiffness is the superfluid density.

• Many-body quantum mechanics 12.09.01. Identical-fermion antisymmetry and the exchange interaction set the framework within which the paired many-body state is constructed.

Historical & philosophical context Master

In 1908 Kamerlingh Onnes liquefied helium at 4.2 K in his Leiden laboratory, opening the low-temperature regime. Three years later he measured mercury's resistance plunging to an immeasurably small value at 4.2 K, the discovery of superconductivity ^{[Kamerlingh Onnes 1911]}. For two decades the phenomenon resisted explanation: perfect conductivity was grasped, but the Meissner effect (1933) was needed to reveal superconductivity as a genuine thermodynamic phase rather than frozen ideal conductivity.

The phenomenological Ginzburg-Landau theory (1950) introduced the complex order parameter and its free-energy functional, predicting flux quantisation before the charge of the condensate carrier was known to be $2e$ ^{[Ginzburg & Landau 1950]}. The microscopic breakthrough came in 1957 with Bardeen, Cooper, and Schrieffer, whose variational ground state, Cooper instability, and energy gap unified the phenomenology under one quantum-mechanical account ^{[Bardeen, Cooper & Schrieffer 1957]}. Josephson's 1962 prediction of coherent pair tunnelling ^{[Josephson 1962]} turned superconductors into precision electrical standards.

Philosophically, superconductivity is the cleanest laboratory example of spontaneous symmetry breaking and of emergent law: a macroscopic, centimetre-scale object behaves as a single quantum wavefunction. The 1986 discovery of cuprate superconductors by Bednorz and Müller ^{[Bednorz & Müller 1986]} broke the BCS paradigm: decades on, the mechanism of high-$T_c$ pairing remains the deepest open problem in condensed-matter physics, a reminder that emergent order can outrun the microscopic theories built to explain it.

Bibliography Master

@book{Tinkham1996,
  author = {Tinkham, Michael},
  title = {Introduction to Superconductivity},
  edition = {2},
  publisher = {McGraw-Hill / Dover},
  year = {1996},
}

@book{deGennes1989,
  author = {de Gennes, P.-G.},
  title = {Superconductivity of Metals and Alloys},
  publisher = {Westview Press},
  year = {1989},
}

@book{Leggett2006,
  author = {Leggett, Anthony J.},
  title = {Quantum Liquids: Bose Condensation and Cooper Pairing in Condensed-Matter Systems},
  publisher = {Oxford University Press},
  year = {2006},
}

@article{BardeenCooperSchrieffer1957,
  author = {Bardeen, J. and Cooper, L. N. and Schrieffer, J. R.},
  title = {Theory of Superconductivity},
  journal = {Physical Review},
  volume = {108},
  pages = {1175--1204},
  year = {1957},
}

@article{KamerlinghOnnes1911,
  author = {Kamerlingh Onnes, H.},
  title = {Further experiments with liquid helium. On the sudden change in the rate at which the resistance of mercury disappears},
  journal = {KNAW Proceedings},
  volume = {13},
  pages = {1274--1283},
  year = {1911},
}

@article{GinzburgLandau1950,
  author = {Ginzburg, V. L. and Landau, L. D.},
  title = {On the theory of superconductivity},
  journal = {Zhurnal Eksperimental'noi i Teoreticheskoi Fiziki},
  volume = {20},
  pages = {1064--1082},
  year = {1950},
}

@article{Josephson1962,
  author = {Josephson, B. D.},
  title = {Possible new effects in superconductive tunnelling},
  journal = {Physics Letters},
  volume = {1},
  pages = {251--253},
  year = {1962},
}

@article{BednorzMuller1986,
  author = {Bednorz, J. G. and M{\"u}ller, K. A.},
  title = {Possible high {$T_c$} superconductivity in the {Ba--La--Cu--O} system},
  journal = {Zeitschrift f{\"u}r Physik B},
  volume = {64},
  pages = {189--193},
  year = {1986},
}